Incident data preprocessing in railway control systems using a rough-set-based approach

Abstract

The paper proposes a novel approach to data preprocessing about incidents in railway control systems. Various classes of railway control systems and requirements to their functioning are described. The possible solution concern forming incident data based on diagnosis information is proposed. The main approach is related to rough sets theory, therefore, the basic ideas and mathematical notations in the rough sets area are presented. As the main result, the incident data preprocessing method based on rough sets membership function is developed. The pseudocode of algorithm and numerical examples also are provided.